June 9, 2005

Does 'community' still exist in open source?

Author: Matt Asay

The nature of the open source community is changing. I'm not exactly sure what "open source community" means anymore. When I first got involved with open source in 1998/99, the community was distinct: It was Eric Raymond, Bruce Perens, Robin Miller, and others like them. Developers. Gear heads. Hackers.
Today, it's unclear whether that community still exists in any separate, discernible form.

Look at their former haunts (Apache and Linux mailing lists, among others) and you may still find them, but many have been hired by Novell, IBM, Hewlett-Packard, or some corporation, including a range of startups interested in cashing in on the open source gold rush. In addition to these hard-core developers, there is a significant, growing crowd of "newbies": casual developers, corporate developers (those who work for Morgan Stanley, Unilever, and the like), and a range of other, interested parties.

Community matters -- but what is the community?

I suppose this is a positive evolution as open source leaves the exclusive realm of the geeks and goes mainstream, but it still seems troubling. Troubling because a rapidly expanding, diffusing community may not operate as efficiently and effectively as it did in the past.

This is not simply a "life was better in the good old days" sentiment, either. It is core to my company's business. A friend told me recently that he was leaving SUSE... forever. SUSE is an operating system, not his wife, so while I was concerned, it wasn't something that required marital therapy or anything drastic.

Still, his reasoning intrigued me. He felt that there wasn't enough community support for SUSE. Or Red Hat. He was leaving the large, commercial distributions for Ubuntu, a newcomer with a strong community focus. In his opinion, the best companies are those that foster communities and then supplement them with add-ons, extensions, etc. Community comes first. The worst are those that want to own everything and let the community nibble at the edges.

While his is just one opinion, and decidedly not the opinion of the large crowd of CIOs rushing to adopt the commercial strains of Linux, including Novell's, it marks a potential problem for vendors. In our rush to commercialize Linux and other open source projects, we tend to cloud the community aspect, which obviates many of the benefits vendors (and customers) derive from open source in the first place. Word-of-mouth marketing, supra-corporate QA testing, etc. These benefits disappear when community is trampled in the rush to commercialize open source.

Is there enough community to share?

This calls to mind a related concern: What if the open source community runs out?

That is, everyone these days seems to be building open source-based businesses. While most don't kid themselves about mythical hordes of developers contributing back to their projects, they still rely on the community to both contribute code and to spread the good word. This worked fine (well, not so fine -- a large number of SourceForge.net projects continue to be hollowed out orphans with no code, no activity, and no future) at first, but will it continue to work when everyone is crowding the space, hoping to exploit open source distribution and development?

Are there enough developers to share around? If so, and related to a concern noted above, will anyone derive benefits from those developers, dispersed and fragmented as this new "community" seems to be?

Think about it in terms of billboard advertising. One billboard might mean lots of eyeballs, and lots of brand recognition. One billboard, standing alone, sits apart. It makes an impression.

Now think about Highway 101 x 10. That's just noise, and ugly noise at that.

If, in fact, bandwagon open source startups and enterprise strategies are stretching an already relatively thin open source community, then we may be muddying that community with "billboards." This expansion of the community might make it harder to coalesce support for open source projects, leaving end users with a morass of choice.

Of course, it may well be that the population of open source developers/onlookers is in its nascent stage; that we're nowhere near saturation given that we're still bringing much of the world online. After all, the open source community tends toward production and, hence, more bodies mean more development, or more productivity.

Fair enough. But in the short term, I'm mildly to moderately concerned that we're going to deafen the still tiny open source community with various commercial and community entities plying their respective projects/products. At that point, it's unclear to me that sales/marketing dollars spent in open source efforts, currently significantly less than in traditional software models, will be able to remain low, making open source as a distribution strategy less efficient than originally envisioned.

Microsoft and the community

All of which brings me to Microsoft. It is the one company that has largely been absent from the open source community, yet arguably has the most to contribute to (and derive benefit from) open source. Linux is competitive with Windows; it does not follow, therefore, that open source is competitive with Microsoft. Open source is a process, not a thing. Last time I checked, Microsoft was reasonably good with process, and certainly had nothing to fear from an efficient development methodology.

I have a proposition to make, one that I've encouraged Jason Matusow to speak about at OSBC Boston later this year (Nov. 1-2, 2005). Windows is arguably the world's most pervasive platform (probably equal to or greater than the Internet at this point, frankly, as many who use Windows don't actively use the 'Net, or do so at dial-up speeds). Why not foster robust open source communities on Windows and Internet Explorer? Not internally baked development but external projects that add value to Windows and IE and, hence, provide value for end users?

The default open source platforms are, of course, Linux and FireFox. This is already happening. Even with closed source software, for example, Firefox is increasingly getting equal time to IE: I don't believe I saw a single demo at this year's DEMO conference that wasn't shown on Firefox, instead of IE. (Of course, all of them also ran on IE, but it was significant that they were demo'd on Firefox Ð IE is "out.") Open source platforms are gaining mindshare -- it's only a matter of time before that mindshare translates into serious market share.

Yes, Windows is still king/queen for now, but I can see this starting to fade... not through the enterprise, which is quite conservative, but rather through the consumer space. Already, many of the applications I willingly, gladly use on my computers (Windows, Mac, and Linux) are open source. Not because they're open source, but because they're better. From Firefox to VLC (media player that can handle anything), my desktop increasingly looks like an open source playground, with more and more of my "office suite" functions happening in my email client. I'm nearly at the point of living a Microsoft-free existence, and I never consciously set out to get there.

If Microsoft wants to forestall defections and improve the user experience for existing customers, it should foster Windows and IE as breeding grounds for open source application innovation. Why concede that ground to Linux, or leave it to happenstance that open source developers will find their way to Windows? This is a pain-free, easy-to-kickstart program for Microsoft that could yield tangible benefits. Open source is not the enemy. It is the opportunity.

If there's room for gargantuan Microsoft in the open source community, and there is, then my concerns about its rapid expansion are probably overblown somewhat. Still, I think new entrants to the community need to focus first on what they can bring, rather than what they can take. So long as it is, on average, a contributing community -- rather than a pilfering community -- it should be big enough for everyone. Including Microsoft.

Matt Asay, a regular contributor to ITMJ, is part of Novell's Linux Business Office and chairs the company's Open Source Review Board, responsible for evangelizing the benefits of open source within and outside Novell, and for helping to generate Novell's open source strategy. He also is the co-founder of the Open Source Business Conference.

Click Here!